ARTICLE pubs.acs.org/est
Enhanced Transport of Colloidal Oil Droplets in Saturated and Unsaturated Sand Columns Micheal J. Travis, Amit Gross, and Noam Weisbrod* Department of Environmental Hydrology & Microbiology, Zuckerberg Institute for Water Research, The Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, Midreshet Ben-Gurion, 84990 Israel
bS Supporting Information ABSTRACT: Colloidal-sized triacylglycerol droplets demonstrated enhanced transport compared to ideal latex colloid spheres in both saturated and unsaturated quartz sand columns. Oil droplets (mean diameter 0.74 ( 0.03 μm, density 0.92 g cm 3, ζ-potential 34 ( 1 mV) were injected simultaneously with latex microsphere colloids (FluoSpheres; density 1.055 g cm 3, diameters 0.02, 0.2, and 1.0 μm, ζ-potentials 16 ( 1, 30 ( 2, and 49 ( 1, respectively) and bromide into natural quartz sand (ζ-potential 63 ( 2 mV) via short-pulse column breakthrough experiments. Tests were conducted under both saturated and unsaturated conditions. Breakthrough of oil droplets preceded bromide and FluoSpheres. Recovery of oil droplets was 20% greater than similarly sized FluoSpheres in the saturated column, and 16% greater in the 0.18 ( 0.01 volumetric water content (VWC) unsaturated column. Higher variability was observed in the 0.14 ( 0.01 VWC column experiments with oil droplet recovery only slightly greater than similarly sized FluoSpheres. The research presents for the first time the direct comparison of colloidal oil droplet transport in porous media with that of other colloids, and demonstrates transport under unsaturated conditions. Based on experimental results and theoretical analyses, we discuss possible mechanisms that lead to the observed enhanced mobility of oil droplets compared to FluoSpheres with similar size and electrostatic properties.
’ INTRODUCTION Triacylglycerol (i.e., food oil) contamination of the environment may occur in various ways such as: soil-based waste disposal of materials from edible oil or other food processing operations;1 bioremediation of recalcitrant organic pollutants in which edible oil is introduced to the subsurface;2,3 and irrigation with wastewater.4 Edible oils are essentially insoluble in water, but may be dispersed in solution as colloidal-sized droplets from a few nanometers to several micrometers in diameter. Emulsion droplets in water may be stabilized by surfactants or finely divided solids,5 and present difficult treatment challenges.6 If solutions containing oil emulsion droplets are applied to, and move through the soil, they can enhance the transport of oil soluble compounds such as pesticides, pharmaceuticals, or environmental estrogens.7 The transport of edible oil emulsion droplets through saturated porous media has been previously published (e.g., refs 2,3,8 11). However, many of these studies focused on relatively large oil droplets (several μm8), or highly concentrated emulsion solutions (e.g., >10% oil3,9). To the best of our knowledge, the direct comparison of dilute oil colloid mobility in porous media to that of latex microsphere “ideal” colloids10 14 has never been reported. Furthermore, the transport of oil emulsion droplets in unsaturated media has not been documented. Transport characteristics of colloids in porous media may be influenced by water content,12 pore velocity,13 colloid r 2011 American Chemical Society
concentration,14 pH,15 and ionic strength.16 Colloid retention is influenced by factors at the interface, collector, and pore scales.17 Filtration theory predicts the transport of colloids through porous media.18 Derjaguin Landau Verwey Overbeck (DLVO) theory quantifies electrostatic energies to predict colloid and collector surfaces interactions.19,20 Colloid colloid and colloid grain surface interactions may be influenced by surface charge heterogeneity,21 the air water interface,22 Lewis acid base interactions,23 and steric contributions.24 The link between colloid transport in unsaturated versus saturated media has been explored mainly for latex microspheres as “ideal” colloids (e.g., refs 17,25 27). Nevertheless, the interactions of physical and chemical processes that govern unsaturated colloid transport and retention are still not well understood.17,28 Multiple interfaces in unsaturated media (e.g., solid solid, solid water, air water, air water solid) increase potential colloid retention through wedging/straining, bridging, or film straining.16,17,29 Furthermore, different colloids have been shown to possess unique transport qualities (e.g., biocolloids30,31 and clay13,32). Received: December 2, 2010 Accepted: September 28, 2011 Revised: September 6, 2011 Published: September 28, 2011 9205
dx.doi.org/10.1021/es104040k | Environ. Sci. Technol. 2011, 45, 9205–9211
Environmental Science & Technology The objective of this research was to determine the transport characteristics of colloidal-sized oil droplets under conditions in saturated and unsaturated natural sand, and to compare with the mobility of “ideal” latex microspheres.
’ MATERIALS AND METHODS Soil Columns and Porous Media. Short-pulse tracer experiments were conducted in soil columns (9.9 cm diameter, 30 cm long) equipped to control and monitor volumetric water content (VWC) and matric pressure. The setup was similar to systems used in previous studies.33,34 A detailed schematic is included in Figure SI-1 of the Supporting Information (SI). The columns were packed with washed and sieved (300 500 μm) natural beach sand from the coastal region of southern Israel, as detailed in SI S1. No clay-minerals or organic matter were detected in the sand. The sand was >99% quartz, with iron and other metals detected as potential oxide coatings by elemental analysis of the grain surfaces (SI Table SI-1). Three separate column packs were used, each for a set of four replicate experiments: (1) saturated; (2) unsaturated “high” water content (0.18 ( 0.01 VWC); and (3) unsaturated “low” water content (0.14 ( 0.01 VWC), so that porosity and pore structure were uniform within each set of experiments. Average pore diameter for the sand was 75 ( 2 μm calculated from capillary rise experiments,35 described in SI S2. Bulk density of the sand was 1.72 ( 0.02 g cm 3, and porosity 0.35 ( 0.007. Saturated hydraulic conductivity, measured by the constant head method, was 0.039 ( 0.001 cm s 1. ζ-potential of the sand grains was 63 ( 2 mV (pH 7), calculated from streaming potential (Anton Paar, Graz, Austria) using the Helmholtz-Smoluchowski equation and the Fairbrother-Mastin approach.36 Soil column and sand characteristics are summarized in SI Table SI-2. Solution was introduced to the upper surface of the column via a rain simulator. Background Solution and Pulse Preparation. All experiments used artificial rainwater (ARW)37 for the background and tracer solutions (SI S3) with ionic strength 0.021 mM, pH 7.2 ( 0.2, and electrical conductivity 183 ( 10 μS cm 1. Tracers included (1) oil droplets (100 mg L 1); (2) 1.0, 0.2, and 0.02 μm diameter FluoSpheres (Invitrogen Corporation, Eugene, OR, at concentrations 1, 5, and 10 mg L 1, respectively); and (3) lithium bromide (40 mg L 1). FluoSpheres are monodisperse, carboxylate-modified latex microspheres impregnated with fluorescent dye. FluoSpheres sizes were selected within the approximate range of oil droplet sizes (described below) in order to enable comparison. Oil droplet preparation is detailed in SI S4. Oil droplets were prepared first in a stable, concentrated emulsion consisting of 45% refined sunflower oil, 2.5% surfactant and 52.5% ARW. The concentrated emulsion was then diluted into the tracer solution. Final surfactant concentration in the tracer solution was 5 mg L 1,
